The relationship between the gut microbiome and human disease is well appreciated but poorly understood. Central to gut microbial community dynamics is how nutrients flow between community members as well as between the community and its host. A presumption exists that nutrient exchange in toto is highly complex and interconnected; as such, functional pathways and circuits remain poorly characterized. The main hypothesis for this proposed program is that nutrient flow between the microbiome and the host is relatively stable despite individual microbiome variation; notwithstanding such stability, there exist central microbial species that drive key metabolic reactions between microbiome and host and that lead to disease when disrupted. With our complementary expertise in metagenomics, metabonomics, computational modeling, and gut microbiome biology, we will address this central hypothesis with the following specific aims: (1) Reconstruct and validate metabolic networks of in vitro synthetic microbial communities directly from meta-transcriptomic sequencing data. We have developed a novel systems biology approach to reconstruct the metabolic networks of individual members of a microbial community directly from meta-transcriptomic data. We will validate this approach with in vitro synthetic microbial communities derived from the altered Schaedler flora; (2) Characterize the in vivo correlation between the ileal microbiome and host metabonome. We will characterize the host (mouse) metabonome (from urine, serum, and ileal lumen) and the ileal microbiome of coupled samples under conditions that alter the microbial community (broad- and narrow-spectrum antibiotics). We will also sequence the meta-transcriptomes of the associated microbiomes. These mappings between the ileal microbiome and host metabonome will be delineated over a longitudinal study with antibiotic-mediated perturbation and subsequent recovery; and (3) Predict and validate stabilizing microbial species within the gut microbiome that account for robustness of the host metabonome and describe corresponding metabolic functionalities. We will predict the keystone species that buffer the metabonome from changes in the microbiome and thus, when disrupted, have the most significant impact on variation in the metabonome. We will characterize the key metabolic pathway niches in the microbiome that enable stability of the host metabonome. We will validate these predictions by selectively killing target species via bacteriophage therapies t manipulate the metabonome as directed by results from the computational modeling and antibiotic-mediated perturbation and recovery studies. The implementation of this proposed program will address fundamental questions in the complex nutrient flow dynamics between host and microbiome. An understanding of the relationship between microbiome composition and host metabolism will be key to the development of probiotic, nutritional, and other therapeutic strategies.